Download Neurophysiological correlates of reward processing and cognitive control in Borderline Personality PDF

TitleNeurophysiological correlates of reward processing and cognitive control in Borderline Personality
LanguageEnglish
File Size4.1 MB
Total Pages283
Document Text Contents
Page 1

Neurophysiological correlates of reward processing and

cognitive control in Borderline Personality Disorder patients

with and without self-harm history



PhD Dissertation presented by:

Daniel Vega Moreno



Supervised by:

Dr. Antoni Rodríguez Fornells

Dr. Rafael Torrubia Beltri





PhD in Psychiatry

Department of Psychiatry and Forensic Medicine





Bellaterra, 2014

Page 2

ii | D a n i e l V e g a

Page 141

141 | D a n i e l V e g a NEUROPHYSIOLOGICAL CORRELATES




College, London, UK, www.fil.ion.ucl.ac.uk/spm/). The two functional runs were first

realigned and a mean image of all the EPIs was created. The T1-weighted image was

coregistered to this mean EPI image and then segmented into grey and white matter

(GM; WM) by means of the New Segment toolbox included with SPM8 (Ashburner &

Friston, 2005). Following segmentation, grey and white matter images were fed to

DARTEL (Ashburner, 2007) in order to achieve normalization. After normalization,

data was resampled to 2x2x2 mm³ and spatially smoothed with an 8x8x8 full width at

half maximum (FWHM) Gaussian kernel.

An event-related design matrix was specified using the canonical hemodynamic

response function. Onsets for each condition were modeled at the moment in which

participants received the feedback. Data were high-pass filtered (to a maximum of 1/128

Hz) and serial autocorrelations were estimated using an autoregressive (AR(1)) model.

Motion effects were minimized by also including in the model the movement

parameters estimated during the realignment phase. First-level contrasts were specified

for all participants using each condition (gain, boost gain, loss, boost loss, blank)

against the implicit baseline. The contrast images from all participants in the three

groups were introduced into a mixed-design analysis of variance (ANOVA) with

condition (gain, boost gain, loss, boost loss, blank) as a within-subjects variable and

with Group (Healthy, NI-BPD, SI-BPD) as a between-subjects variable. A general gain

(gain and boost gain) > loss (loss and boost loss) contrast for all groups (Healthy, NI-

BPD and SI-BPD) was calculated to show the expected activations in reward-related

areas (Camara et al., 2010). In addition, for each condition (gain, boost gain, loss, boost

loss, blank) and for gain>loss and boost gain> boost loss the effect of Group was

assessed with an F-test. Finally, for each condition showing a reliable Group effect,

Page 142

142 | D a n i e l V e g a NEUROPHYSIOLOGICAL CORRELATES




two-sample t-tests (NI-BPD > Healthy, SI-BPD > Healthy, SI-BPD > NI-BPD and their

respective reversed contrasts) were planned, to check for the direction of the effect.

All statistics in Figures and Tables are reported at a p < 0.001 uncorrected

threshold with a minimal cluster size of 20 voxels (Lieberman & Cunningham, 2009).

Peaks surviving a p<0.05 FWE-corrected threshold are indicated in tables. Maxima are

reported in MNI coordinates. Anatomical and cytoarchitectonical areas were identified

using the Automated Anatomical Label atlas (Tzourio-Mazoyer et al., 2002) and the

Talairach Daemon database atlases (Lancaster et al., 2000) included in the xjView

toolbox (http://www.alivelearn.net/xjview8/).

Functional-connectivity analysis. An exploratory connectivity analysis was also

performed. An 8 mm radius ROI was defined around the peak value in left Orbitofrontal

Cortex (OFC; -32 58 -14) of the F-test showing a Group effect for the boost gain

condition (the only condition showing a significant Group effect, see Results below).

Individual time-courses from this ROI were extracted, and an extended model was

created, including the five conditions previously defined (gain, boost gain, loss, boost

loss, blank) plus the extracted OFC time-course and the derived psychophysiological

interaction (PPI) within the standard PPI approach (Friston et al., 1997) as regressors.

PPIs were used to test for higher inter-regional coupling with the OFC during boost

gains. Second level independent t-tests (NI-BPD > Healthy, SI-BPD > Healthy, SI-BPD

> NI-BPD and the reversed contrasts) were computed.

For this exploratory connectivity analysis, a more lenient p < 0.005 uncorrected

threshold with a minimal cluster size of 20 voxels, was used (Lieberman &

Cunningham, 2009). Maxima and all coordinates are reported in MNI coordinates.

Page 282

282 | D a n i e l V e g a NEUROPHYSIOLOGICAL CORRELATES




Abstract

Background: The psychological profile of Borderline Personality Disorder (BPD)

patients, with impulsivity and emotional dysregulation as core symptoms, has guided

the search for abnormalities in specific brain areas such as the hippocampal-amygdala

complex and the fronto-medial cortex. So far, though, whole brain imaging studies have

delivered highly heterogeneous results involving different brain locations.

Methods: Resting state functional Magnetic Resonance Imaging (MRI) and diffusion

MRI was acquired on BPD patients and on an equal number of matched controls (N =

60 for resting and N = 43 for diffusion). While Mean diffusivity (MD) and Fractional

Anisotropy (FA) brain images were generated from diffusion data, the Amplitude of

Low Frequency Fluctuations (ALFF) and Global Brain Connectivity (GBC) images

were used for the first time to evaluate BPD related brain abnormalities from resting

functional acquisitions.

Results: Whole brain analyses using a p = 0.05 corrected threshold showed a

convergence of BPD alterations in genual and perigenual structures, with frontal white

matter FA abnormalities partially encircling clusters of increased MD and GBC values.

A cluster of enlarged ALFF (high resting activity) was located in part of the left

hippocampus and amygdala. In turn, this cluster showed increased resting functional

connectivity with a cluster in the anterior cingulate.

Conclusions: With a multimodal approach, and without using a priori selected regions,

we prove that structural and functional abnormality in BPD involves both temporo-

limbic and fronto-medial structures, as well as their connectivity, all of them

extensively related to behavioral and clinical symptoms in BPD patients.

Similer Documents